Distinguished Machine Learning Engineer and Data Scientist specializing in scalable AI systemsAdvanced ML & Deep Learning: Production implementation of Large Language Models (LLaMA3.1, Mistral-7B, BERT), Neural Networks (CNN, RNN, GANs), Time Series models (Prophet, ARIMA, SARIMA), Vision Transformers, and advanced architectures (ResNet, LSTM). Deep expertise in RAG (Retrieval-Augmented Generation), model ensembling, transfer learning, and fine-tuning. Advanced statistical modeling including Bayesian inference, GLMs, and survival analysis. Good experience in XGBoost, LightGBM, CatBoost optimization.Data Science: - Advanced feature engineering and automated feature selection- Clustering (DBSCAN, HDBSCAN, K-means++) and anomaly detection- Deep reinforcement learning for optimization problems- Natural Language Processing (topic modeling, sentiment analysis)- Time series forecasting and sequential pattern mining- Recommendation systems (collaborative filtering, content-based)- Dimensionality reduction (PCA, t-SNE, UMAP)- Causal inference and A/B testing design Model Development & Optimization:- Custom loss functions for domain-specific problems- Advanced hyperparameter optimization (Bayesian, grid, random, etc)- Model distillation and quantization techniques- Cross-validation strategies for complex data splits- Handling imbalanced datasets (SMOTE, ADASYN)- Ensemble methods (stacking, blending, voting) Data Engineering: - Production-grade data pipelines and feature stores- Statistical analysis and hypothesis testing- Experimentation frameworks and metric design- Data preprocessing pipelinesTechnical Stack:- ML: PyTorch, TensorFlow, Scikit-learn, Rapids- Data: Pandas, NumPy, SciPy, Polars, Dask- Visualization: Plotly, Seaborn, Bokeh- Cloud: AWS SageMaker, Vertex AI, AzureML- MLOps: MLflow, Weights & Biases, DVC Key Projects:> Large-scale Recommendation Engine in caselaw, healthcare and financial sectors:- Multi-stage ranking system with embedding-based retrieval- Custom loss functions for ranking optimization- Advanced feature crossing and interaction modeling- Multi-objective optimization for business metrics> Advanced Analytics Platform:- Automated anomaly detection system- Time series forecasting with uncertainty- Causal impact analysis framework- Segmentation using deep learningMLOps:- Automated model retraining pipelines on kubernetes using argo workflows and kuberay- Model monitoring and drift detection- Scalable inference optimization- Multi-region model deployment- Experiment tracking and versioning